550 REPOET OF COMMISSIONER OF FISH AND FISHERIES. [!)fij 



cleavage. In fact, it gives us a clew to wliy it is that tbere is sncli a 

 distinction between ova as the evenly and unevenly segmenting or the 

 holoblastic and meroblastic types. It will also be seen that it is prob- 

 able that the two types pass into each other by inseusible gradations, 

 which we find is truly the fact when we come to institute a large series 

 of comjiarisons between observed types of development. Moreover, 

 the food-yelk has simply a physiological significance; it is merely a 

 store of food, which has been suj)eradded during the intraovariau 

 growth and maturation of the egg, yet the effect of this superadded 

 yelk is to modify the process of segmentation ijrofoundly. That store 

 of deutoplasm which is added to the egg to nourish the embryo and 

 carry it to that condition of development when it can in a measure take 

 care of itself, is also of i)rofound significance in relation to natural selec- 

 tion, by the operation of which it can alone be supposed that yelks 

 were evolved. This all-comprehending Darwinian law is therefore seen 

 to have influenced the mode of segmentation of ova hy and through the 

 minor and secondary law of nuclear displacement which has been indi- 

 cated above. 



The degree also to which the nucleus is transported from its primi- 

 tively central position determines the degree of inequality of the first 

 segmentation. It is now in the highest degree probable that in the for- 

 mation of the germinal disks of meroblastic ova the process is one and 

 the same throughout the animal kingdom, viz, that its development is 

 accomplislied by the migration and concentration of tlie germinal mat- 

 ter of the egg at its animal pole. We have evidence in superabundance 

 in favor of such a view of the matter, in a great many departments of 

 embryology. There is an evident tendency on the part of the germinal 

 protoplasm of ova to separate itself spontaneously from the food yelk or 

 deutoplasm and aggregate itself either superficially' over the whole 

 surface of the ovum, as in the case of centrolecithal, or at one pole mainly, 

 as in meroblastic or telolecithal types. We will find, however, that the 

 distinction between these two forms is primarily less important than might 

 be supposed, for the meroblastic type passes through a distinct centro- 

 lecithal stage prior to the development of the genninal disk in Gadus, 

 while in other forms the mode of disk development is complicated by a 

 network passing down into and between the deutoi)lasm masses from 

 the external stratum of germinal protoplasm, as in the ova of Clupeoids 

 and Leuciscus. The remarkable centrolecithal segmentation and ar- 

 rangement of the protoplasm of the eggs of the arthropods is, therefore, 

 found not to be so radically ditferent from the usual ty])e as might at 

 first be supposed. Its arrangement, under slightly different laws of 

 segmentation, is referable to the same fundamental principle governing 

 the dissociation and aggregation of the protoplasm and deutoplasm into 

 separate masses. I would also regard the deutoplasm as almost entirely 

 passive in the process of its absorption during the later stages, for we 

 have seen that it is actually ai)propriated by a remnant of the original 



